The purpose of this research is to contribute to the reduction of the adverse outcomes of pregnancy in low income women through the identification of modifiable psychosocial risk factors as predictors. The specific aims of the proposed study are to examine: (1) the relationships between the attachments of low income pregnant women with mother/surrogate and husband (partner); (2) the relative strength of their current attachments with mother/surrogate and husbanD (partner), other forms of social support, life stress, and anxiety as predictors of low income women's psychological well-being; (3) the impact of psychosocial variables, psychological well-being and negative health behaviors (smoking and drug use? on pregnancy outcomes of low income women. A convenience sample of 370 women, 185 each of Afro American and Euro American/Hispanics, 18 through 35 years of age, medicaid eligible, between 14 and 22 weeks of pregnancy, without preexisting medical conditions, and fluent in English, will be utilized for this prospective correlational study. Access to subjects will be through obstetric clinics in the city. Self report questionnaire to measure attachments, social support, life stress, anxiety, and negative health behaviors (smoking, drug use), psychological well-being, will be administered individually, between 14 and 22 weeks, and between 32 and 42 weeks of pregnancy. The hospital records of women, who participate, will be reviewed by-the investigator/research assistant to determine pregnancy, labor, delivery, and newborn complications. The time for completion of the questionnaire will vary from 30-50 minutes. The subjects will be paid $5.00 after each testing for partial compensation for parking and child care. Data analysis will include descriptive statistics, correlational and multiple regression; paired tests, and discriminant analyses. Subjects will be stratified into two ethnic subgroups to test the study hypotheses. Separate analysis will be performed for African American and Euro American/English-speaking Hispanic groups in order to control for the effect of ethnicity. If incidence of complications and magnitude and direction of effect does not differ, subjects will be pooled for further analysis.